You are here
Tech-savvy residents go nimby on self-driving cars
KAREN Brenchley is a computer scientist with expertise in training artificial intelligence, but this longtime Silicon Valley resident has pangs of anxiety whenever she sees Waymo self-driving cars manoeuvre the streets near her home.
The former product manager, who has worked for Microsoft and Hewlett-Packard, wonders how engineers could teach the robocars operating on her tree-lined streets to make snap decisions, speed and slow with the flow of traffic and yield to pedestrians coming from the nearby park. She has asked her husband, an award-winning science-fiction author who does not drive, to wear a shiny vest while cycling to ensure autonomous vehicles spot him in a rush of activity.
The problem is not that she does not understand the technology. It is that she does, and she knows how flawed nascent technology can be.
"I'm not sceptical long-term," said Ms Brenchley, who has lived in Silicon Valley for 30 years. "I don't want to be the guinea pig. I don't want my husband to be the guinea pig."
Ms Brenchley and others who live among the world's technology giants represent a surprising Silicon Valley paradox: residents believe in the power of technology to change the world for the better, but they are sceptical of the role that it might play in their daily lives. This is especially visible as driverless cars from numerous tech giants arrive en masse in the streets of Silicon Valley neighbourhoods.
Some residents say they are confident that the technology can work in limited settings, such as test tracks or simulations. But the software that controls the cars needs to be trained on real-life situations: left-hand turns, bikers, children running out into the streets. And, some residents said, that brings a form of disruption that will tangibly change the fabric of their communities and could even prove dangerous. That became apparent last year, when an Uber robocar struck and killed a pedestrian in Arizona.
Waymo spokeswoman Alexis Georgeson said that safety is core and the company's highest priority, and that the technology could make roads safer. The company's employees and families work and live there, and test the vehicles, too. It is also educating the public at local events. "Our vehicles are programmed to be safe and cautious drivers," she added in a statement.
Sceptical of advanced technology
Silicon Valley types can be most sceptical of advanced technology because they know how it works and what its risks are. Parents with experience at large tech firms have famously cracked down on screen time for their children. Some tech executives will not let female family members ride alone at night with ride-sharing cars. Others keep their kids off social media indefinitely.
That same scepticism has landed on Silicon Valley streets. George Azzari, 39, spent five years in the Mountain View neighbourhood on the doorstep of Google's self-driving subsidiary Waymo. He said the cars tended to form a trail down a small road near his home at rush hour, clogging up traffic. "I basically would run into a bunch of these cars coming back to headquarters at 25 miles per hour (72 kmh) on this tiny road. You can't pass them," said Mr Azzari, chief technology officer at Palo Alto-based social impact start-up Atlas AI. "I definitely drive different when I'm around those things," coming to a hard stop or pulling around one driving particularly slowly.
California has awarded permits to 63 different companies to test self-driving vehicles on state roads, according to state figures from Aug 9. Among them are a slew of tech companies with a substantial Silicon Valley presence: Lyft, Tesla, Alphabet-owned Waymo, General Motors' Cruise division, Ford-affiliated Argo AI, and start-ups such as Aurora and Zoox.
The companies outfit their autonomous cars with complex sensors such as radars and cameras. They are frequently equipped with a cone-like lidar sensor atop a roof-mounted contraption that looks like an upside-down sled. Most are small SUVs or vans that stop and start with regular traffic - driving tasks normally left to the human brain. Safety drivers are in the vehicles to monitor the cars' performance.
While much of the testing is done on closed courses imitating city streets or virtual simulations, real roads are essential in teaching the cars' artificial intelligence the myriad real-life situations it could encounter, the companies said.
That happens in Silicon Valley, one of the early homes for such testing. Companies have also started rolling it out in other states where that is allowed, including Arizona, Nevada and Pennsylvania, to add a diversity of weather, road conditions and driving cultures.
Robocars have become a common sight in the tree-lined suburban streets in Silicon Valley, dotted with California's iconic mid-century Eichler-style homes, where upper-middle-class engineers and tech workers reside and where their children roam on bicycles and skateboards.
Some residents are proponents - or at least indifferent - to the autonomous cars on their streets. And plenty of people here are quick adopters on tech, demonstrated in part by the sheer volume of Teslas on the streets. Plus, some of the communities where the testing is taking place say that so far it has largely been incident free, with few complaints from residents.
Brad Templeton, who lives in testing hotspot Sunnyvale, frequently sees the cars on the road. He worked on them, too, as part of Google's self-driving car project roughly a decade ago. Most experts in the field said real-world testing is needed, he noted, something he agrees with.
Mr Templeton said that a small number of crashes is acceptable when considering the eventual overall improved safety when human drivers are off the roads. He compared it to teenagers learning to drive.
"We accept them driving, with very high risk, because it is the only way to turn them into safer middle-aged drivers. And all we get out of that is one safer driver," he said. As autonomous vehicles are trained, "we get a million safer cars from a prototype fleet of hundreds".
More questions than confidence
But John Joss, 85, did not think that the robot drivers are that mature. "They drive like either geriatrics or 17-year-olds who have very limited experience of driving," said Mr Joss, a magazine writer. He said the behaviour of hundreds of vehicles on his Mountain View streets has inspired more questions than confidence. In one instance, he said, a Waymo vehicle drove from 13 kmh to 16 kmh on a 40 kmh street, gathering a "tail" of other vehicles behind, trapping vehicles at a light. In another instance, he said, a Waymo vehicle stacked cars behind a blind curve after struggling to navigate an expressway on-ramp.
Knowing tech workers well has added to his concerns about the risks engineers might be taking on his roads, he said. "I have met, dealt with, interviewed and written about geeks for the last 30 or 40 years," Mr Joss said. While they understand the tech that they are working on, it does not always translate to a broader understanding of potential impact.
Others here see the testing as a nakedly self-interested push for profit, an arms race to launch the first driverless taxi service and reap the profits that come from being unburdened to paying drivers. They do not want to be the unwitting subjects of those tests.
Sally Applin lives in Silicon Valley and frequently ends up on the road with robocars. She said she tries to avoid them when she does. She recently took a visitor around the Valley, who started to take pictures when he saw a Waymo self-driving vehicle. She slowed down to stay safe. "I keep a distance," she said.
A review of safety reports in California found self-driving vehicles are at risk of rear-end collisions. Experts said the vehicles behave unlike human drivers and are hard for people behind the wheel to predict, and they jolt to a stop when detecting a hazard.
As Ms Applin recently sat at a red light, she realised that a driver next to her was going to make an dangerous illegal turn. "How does an autonomous vehicle sense that?" she asked.
Ms Applin, who studies the intersection of people, algorithms and ethics as a researcher but who also feels the impact in her real life, said the introduction of these cars to the road is problematic because it means that she and other drivers are responsible when it comes to training the new tech. WP